Although there have been numerous demonstrations of intact or relatively intact skill learning in amnesics, there have been few attempts to provide a detailed analysis of the nature of such learning or to generate specific hypotheses about when amnesics will show normal skill learning and when they will show impaired skill learning. We propose to accomplish both of these objectives within the context of a process approach to skill acquisition. Specifically, we will examine the role of reactivation in amnesic and normal skill learning. The proposed research should further our understanding of both normal and amnesic skill learning.